replace a string upto and including the first instance of another string in OS X

Question

I'm trying to replace a string upto the first instance of a second string with sed (in OS X). My sed command unfortunately replaces everything up to the last instance of the second string.

my text:

<li>lorem ipsum</li><li>dolor sit amet</li><li>something</li><li></li>

I need to remove <li>lorem ipsum</li> so that the new line looks like this:

<li>dolor sit amet</li><li>something</li><li></li>

my sed command unfortunately replaces the whole line:

sed -i "" 's:<li>lorem.*</li>::'

Do you have an idea how to solve this with sed?

Thanks AleV

Answer 1

I think the closest to what you want is this:

sed -i "" 's:<li>lorem[^<]*</li>::'

(pattern means include anything that is not a "<", i.e. up to the next tag). With sed you'll have to decide one way or another for a terminating character or a set of valid characters.

.

The cleanest solution however would be of course be a non-greedy match, but sed does not offer that (see non greedy regexp) but using perl would be a solution according to the link:

perl -pe 's:<li>lorem.*?</li>::'